Applied Clay Science 108 (2015) 28–39

Contents lists available at ScienceDirect

Applied Clay Science

journal homepage: www.elsevier.com/locate/clay

Research paper

Estimation of the heat of reaction in traditional ceramic compositions

S. Ferrer ⁎, A. Mezquita, M.P. Gomez-Tena, C. Machi, E. Monfort
Instituto de Tecnología Cerámica, Asociación de Investigación de las Industrias Cerámicas, Universitat Jaume I, Castellón, Spain

a r t i c l e i n f o a b s t r a c t

Article history: Most energy studies on the traditional ceramic manufacturing process focus on the firing stage because this is the
Received 16 September 2014 process stage that consumes the greatest amount of thermal energy. At present in Europe, using typical technol-
Received in revised form 12 February 2015 ogies, about 50% of the energy input in the firing stage is still lost through the kiln stacks. A key issue in energy
Accepted 15 February 2015
studies is the accurate determination of the energy efficiency of the process, an issue that may become crucial
Available online 6 March 2015
in coming years to enable the energy management of different facilities and products to be compared. To reliably
Keywords:
determine energy efficiency, accurate determination is required of the energy needed for the necessary physico-
Heat of reaction chemical transformations to develop in the material in the firing stage. This energy is also the only strictly unre-
Physico-chemical transformations coverable energy, as the energy contained in other streams could, potentially at least, be partly recoverable. The
Energy efficiency present study was undertaken to develop an analytical methodology for estimating the heat of reaction of seven
Firing different traditional ceramic products, involving a broad spectrum of compositions, with peak firing tempera-
Traditional ceramics tures ranging from 850 °C to 1200 °C. The following industrial ceramic compositions were studied: four ceramic
tile compositions (red-body stoneware tile, porcelain tile, red-body earthenware wall tile, and white-body earth-
enware wall tile); two structural ceramic compositions (white brick and roof tile), and a porcelain tableware
composition. To estimate the energy involved in the physico-chemical reactions in the firing stage, an analytical
methodology was developed, based on the mineralogy data of the unfired body composition and on the enthalpy
of formation of the minerals in the fired tiles. The methodology was validated by comparing the results with
experimental data.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction In traditional ceramic manufacture, from an energy viewpoint, the

At present in Europe, using typical technologies, large quantities of
thermal energy are consumed in traditional ceramic manufacturing
processes. These processes are quite similar and consist of stages that
differ as a function of the product made. The main production stages
are as follows: raw material preparation (by dry or wet milling),
forming (by casting, pressing, or extrusion), drying, firing of the body
(in double-fired products), glazing and decorating, and firing
(Agrafiotis and Tsoutsos, 2001; Bovea et al., 2010; IPTS. European Com-
mission, 2007).
Thermal energy consumption takes place mainly in three process
stages: wet milling (in which the ceramic slurry may be needed to be
dried for semi-dry pressing); drying (after forming); and firing. In
this manufacturing process, firing accounts for more than 50% and
70% of the overall energy consumption when wet milling and dry
milling are used, respectively (Mezquita et al., 2009, 2014; Monfort
et al., 2010).
⁎ Corresponding author at: ITC, Campus Riu Sec, 12006 Castellón, Spain. Tel.: +34 964
34 24 24; fax: +34 964 34 24 25.
E-mail address: salvador.ferrer@itc.uji.es (S. Ferrer).
firing stage is the most important process stage, as it needs to supply
sufficient heat for the necessary physico-chemical transformations in
the material to develop and to provide the product with the desired
technical and aesthetic properties.
The thermal energy needed to fire glazed ceramics usually stems
from the combustion of natural gas. However, in the manufacture of
structural ceramics, which are mainly unglazed products, other fuels
may be used, such as pet coke, coal, or biomass. These fossil fuels give
rise to air emissions of CO2, a greenhouse gas, and the emissions of
which are subject to international control and mitigation measures.
The present study was undertaken to determine the quantity of
energy consumed by the chemical reactions that develop in the ceramic
firing stage, i.e. in the heat of reaction. At present, using today's technol-
ogies, the heat of reaction constitutes a relatively small fraction (5–20%)
of the overall energy required in industrial firing. It is the only energy
that cannot be recovered, in contrast for example to the sensible heat
of fired ceramics (which is partially recoverable in the cooling phase)
or to combustion gas heat (recoverable using heat exchangers).
From an energy viewpoint, in the firing cycle, the material un-
dergoes both exothermic and endothermic transformations. These
need to take place gradually and in a controlled way, as they could
otherwise lead to permanent defects in the end product.

http://dx.doi.org/10.1016/j.clay.2015.02.019
0169-1317/© 2015 Elsevier B.V. All rights reserved.
 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39 29

1200

Gass transition
1000 range
Sintering

Firing temperature, ºC
Carbonate Liquid-phase formation
decomposition Crystallisation
800 Densification
β-α SiO2 inversion

600 α-β SiO2 inversion

400 Clay mineral dehydroxylation

Combustion of organic matter

200

Loss of water
0
Time, min

Fig. 1. Typical firing curve with the physico-chemical reactions that develop in traditional ceramic compositions.

The shape of the firing curve (temperature as a function of time) constrained by quartz phase transformation at 573 °C, as this pro-
essentially depends on the body composition and, in glazed ceramics, duces a volume change that can cause cracking in the product if
on the nature of the glaze (glossy, opaque, matt, etc.). In general, cooling does not take place uniformly throughout the material,
every traditional ceramic firing cycle includes the following steps: owing to the low thermal conductivity of ceramic materials.

• Heating: In this step, the unfired products are heated from ambient
temperature to about 800 °C (depending on the raw material compo-
sition). This is the step in which outgassing of the ceramic body needs
to take place to avoid problems of bloating, bubbles, pinholing, glaze
porosity, and colour differences at higher temperatures. Various
transformations take place in the material during heating, such as
the removal of free water, combustion of organic matter, allotropic
transformation of α-quartz to β-quartz, loss of OH− groups in the
clays, and carbonate decomposition when the composition contains
carbonates. All these chemical reactions lead to a series of internal
stresses in the ceramic product owing to gas release, changes in vol-
ume, etc. If these stresses exceed the mechanical strength of the as
yet “unfired” ceramic product, failure will occur.
• Firing: The actual firing step runs from about 800 °C to the pro-
grammed peak temperature, typically between 850 °C and 1350 °C,
which depends on the product being made. The main physico-
chemical transformations take place in this step. These reduce the
ceramic material's porosity. They lead to the most important dimen-
sional changes (shrinkage) in low porosity products (stoneware
and porcelain ware) and to the formation of stable crystalline phases
(calcium silicates) in porous products (earthenware wall tiles and
structural ceramics).
• Cooling: This begins when the heat input ends. In the cooling step,
product temperature decreases from peak temperature to near ambi-
ent temperature. In traditional ceramics the cooling rate is essentially
A typical traditional ceramic firing cycle is shown in Fig. 1. The figure
summarises the most important physico-chemical changes that develop
during traditional ceramic thermal treatment and the temperature
ranges in which these changes occur (Pennisi, 1991). A ceramic firing
schedule depends on several factors, the most important being body
composition, thickness and geometry of the ceramic product, kiln load-
ing density, and type of kiln. For instance, the schedule time in roller
hearth kilns (low product thickness and low kiln loading density) is typ-
ically between 30 and 60 min, whereas in tunnel kilns (high loading
density and thick products), this ranges from 18 to 24 h.
The raw material composition plays a very important role, not only
in environmental emissions in the firing stage, but also in energy con-
sumption in this stage. The quantities of carbonates, organic matter,
and other raw materials in the ceramic compositions influence kiln
thermal energy consumption, as the energy required to develop the
physico-chemical reactions in firing depends on the ceramic composi-
tion involved (Mezquita et al., 2009).
Although numerous studies have been conducted on the firing of
traditional ceramics (Bernardo et al., 2010; Castelein et al., 2001; Eliche
et al., 2011; Macgee, 1926; Madivate et al., 2004; Njoya et al., 2012;
Plante et al., 2009; Sánchez et al., 2001; Simpson, 1927; Vogt and Thom-
as, 2012), mainly addressing technical, environmental, or energy issues,
few studies have focused on the energy involved in the physico-
chemical transformations that take place (Macgee, 1926; Simpson,
1927; Madivate et al., 2004; Vogt and Thomas, 2012).

Table 1
Main processing details and end product characteristics of the studied compositions.

Reference Product Processing details End product

Milling Shaping Firing

Peak temperature (°C) Total cycle (h) Kiln Glazed Water absorption (%)

RBT Red-body stoneware tile Wet Pressing 1143 0.46 Roller Yes 2.9
PCT Porcelain tile Wet Pressing 1190 0.57 Roller Yes b0.5
RWT Red-body earthenware wall tile Wet Pressing 1126 0.40 Roller Yes 14
WWT White-body earthenware wall tile Wet Pressing 1140 0.51 Roller Yes 17
WBR White brick Dry Extrusion 850 24 Tunnel No 20
RFT Roof tile Dry Pressing 1019 18 Tunnel No 10
PTW Porcelain tableware Wet Slip casting 1000 22 Tunnel Yes b0.5
30 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39

Table 2 indicates that, at temperatures above 1250 °C, the error becomes
Chemical analysis and loss on ignition (LOI) (%) of the seven test compositions (% by unacceptable (N10%).
weight).

Oxides Composition 3. Materials used and experimental procedure

RBT PCT RWT WWT WBR RFT PTW
3.1. Material characterisation
SiO2 72.02 73.84 65.63 72.12 55.74 63.54 67.05
Al2O3 16.91 19.59 17.21 16.46 19.99 12.44 29.84
Fe2O3 3.79 – 4.11 1.24 1.98 – 0.40 3.1.1. End product characteristics
TiO2 – – – 0.71 – – 0.12 The study was conducted on seven traditional industrial ceramic
CaO 1.50 0.40 7.10 7.15 12.62 18.66 0.18
compositions: four ceramic tile bodies, two structural ceramic bodies,
MgO 1.30 – 1.63 0.23 4.62 0.96 0.15
K2O 2.36 0.97 2.72 1.95 3.70 2.25 0.80
and a porcelain tableware composition. The ceramic tile and tableware
Na2O 2.13 5.20 1.61 0.15 1.36 2.15 1.46 compositions were prepared industrially by wet milling and spray dry-
LOI 5.21 3.88 9.11 9.10 13.44 13.55 8.60 ing the resulting suspensions, whereas the brick and roof tile composi-
Total 100 100 100 100 100 100 100 tions were prepared by dry milling. The ceramic tile compositions were
fired in industrial roller hearth kilns; the other compositions were fired
in industrial tunnel kilns. Only the ceramic tiles and tableware were
glazed. The main processing details and end product characteristics
are summarised in Table 1.
This study was undertaken to develop an analytical methodology for
estimating the heat of reaction in traditional ceramic compositions. To 3.1.2. Chemical analysis
validate the methodology, differential scanning calorimetry (DSC), an The chemical composition was determined by wavelength
experimental thermal analysis technique, was used, which is briefly dispersive X-ray fluorescence spectrometry (WD-XRF), using a
described below. To perform the relevant calculations, thermal data PANALYTICAL model Axios WD-XRF spectrometer with Rh tube and
(such as the calorific capacities and enthalpies of formation) were 4 kW power. Reference materials were used to guarantee measurement
required, on which very little information is at present available for traceability.
these traditional ceramic compositions. The samples were prepared as fused beads and pressed powder
pellets. To prepare the fused beads, the sample was mixed with a
2. Objective and scope 50:50 mixture of LiBO2/Li2B4O7 as flux in a Pt/Au crucible, adding a solu-
tion of LiI as bead-releasing agent. The mixture was then fused using a
In the literature surveyed, no clearly defined methodology was PHILIPS model Perl'X3 automatic fusion bead preparation machine.
found for determining the heat of transformation in traditional industrial The pressed powder pellets were prepared by using a solution of
ceramic compositions (Vogt and Thomas, 2012). n-butyl methacrylate in acetone as binder and they were formed in a
The present study was undertaken with the following objectives: CASMON hydraulic press.
• To define and compare two different methodologies for calculating
3.1.3. Mineralogical analysis
the energy needed to develop the relevant physico-chemical transfor-
The mineralogical composition was determined out by X-ray
mations in ceramic raw material compositions during the firing stage.
diffraction (Brindley and Brown, 1984; Klein and Hurlbut, 1985). The
• To analyse the heats of transformation obtained for seven types of tra-
diffractograms of the powder samples were obtained using a Bruker
ditional ceramic compositions.
D8 Advance theta–theta diffractometer model with copper radiation
The studied ceramic compositions cover a wide variety of traditional (λ = 1.54183 Å Ka). The equipment has a secondary monochromator
products that are fired at peak temperatures ranging from 850 °C to and measurements were made with a Bruker VANTEC solid state detector.
1200 °C. Although the maximum theoretical temperature of the DSC The experimental measurement conditions were 5 to 90° (2θ) with a step
instrument used in the experiments was 1350 °C (as set out below size of 0.015 and 1.2 s/step. The crystalline phases were quantified by the
in point 3.2.1), the study was limited to compositions fired at peak Rietveld method (Zevin and Kimmel, 1995). All diffractograms were
temperatures below 1250 °C, since the authors' experience in this field treated with TOPAS programme version 4.2 (Bruker).

Table 3
Mineralogical analysis and organic matter content of the seven test compositions (% by weight).

Crystalline phases Composition

RBT PCT RWT WWT WBR RFT PTW

Clay minerals Kaolinite 15 ± 2 19 ± 2 13 ± 2 18 ± 2 6±2 4±2 50 ± 2

Illite/muscovite mica 19 ± 2 8±2 21 ± 2 13 ± 2 23 ± 2 13 ± 2 10 ± 2
Chlorite 4±1 – 5±1 – 3±1 – –
Paragonite – – – – 6±1 – –
Carbonates Calcite 2±1 – 11 ± 1 11 ± 1 10 ± 1 21 ± 1 –
Dolomite 1 ± 0.5 – 1 ± 0.5 – 12 ± 1 3±1 –
Feldspars Potassium feldspar 12 ± 1 6±1 8 ± 0.5 7±1 3±1 6±1 4±1
Albite – 36 ± 1 1 ± 0.5 – – 4±1 6±1
Oxides Quartz 44 ± 3 31 ± 3 37 ± 3 51 ± 3 22 ± 3 29 ± 3 30 ± 3
Haematite 3 ± 0.5 – 3 ± 0.5 – 1 ± 0.5 – –
Amorphous phase – – – – – 13 ± 4 20 ± 4 –
Organic matter Biomass (olive pit) – – – – 0.7 ± 0.05 – –
Polyvinyl alcohol (PVA) – – – – – – 0.61 ± 0.05
Polyvinyl glycol (PEG) – – – – – – 0.20 ± 0.05
Total 100 100 100 100 100 100 100.81
 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39 31

Table 4
Mineralogical composition of the fired compositions (% by weight).

Composition

Crystalline phases RBT PCT RWT WWT WBR RFT PTW

Quartz 35 ± 3 24 ± 3 28 ± 3 41 ± 3 19 ± 3 25 ± 3 30 ± 3
Haematite 3 ± 0.5 – 2 ± 0.5 – 1 ± 0.5 – –
Anorthite – – 26 ± 2 18 ± 2 – – –
Wollastonite – – – 3±1 – 5±1 –
Enstatite – – 1 ± 0.5 – – – –
Diopside – – 4 ± 0.5 – 4 ± 0.5 4 ± 0.5 –
Gehlenite – – – 1 ± 0.5 6 ± 0.5 –
Mullite 9±1 8±1 – 4±1 – –
Potassium feldspar 6 ± 0.5 – – – 3 ± 0.5 6 ± 0.5 4 ± 0.5
Albite – 6 ± 0.5 – – 3 ± 0.5 6 ± 0.5
Illite/muscovite mica – – – – 7 ± 0.5 – –
Akermanite – – – – 5 ± 0.5 – –
Periclase – – – – 1 ± 0.5 – –
Amorphous phase 47 ± 4 62 ± 4 39 ± 4 33 ± 4 60 ± 4 51 ± 4 60 ± 4
Total 100 100 100 100 100 100 100

Table 5
Information required for the two methods.

Method Information required Test or source Designation

1 Direct method (direct obtainment of the heat of reaction) DSC Experimental method
2 Determination and quantification of the mineralogical composition X-ray diffraction Analytical method
of the (unfired and fired) ceramic composition
Enthalpies of formation of the compositional constituents – Thermodynamic tables
− Literature sources

The crystalline phases of the powder samples were identified from
the positions and intensities of the diffracted peaks using the ICDD data-
base powder diffraction files (Retrieved January 16, 2014). The solved
structures used to carry out the quantification in the study were ob-
tained from the ICDD PDF 4 database of inorganic structures. The
amorphous phase was determined by adding a known mass concen-
tration of fluorite as internal standard to the powdered samples. The
difference between the weight percentage of the fluorite addition
and quantity determined by the Rietveld method allowed estimation
of the amorphous phase.

Table 6
Reactions considered during thermal treatment and enthalpies of reaction calculated from the enthalpies of formation.

Mineralogical phases and components Reactions Enthalpy

Symbol Value (kJ/kg component i)

Free water r1 H2O(l) → H2O(g) Δh1 2500

Clay mineral dehydroxylation

Kaolinite r2 Al2O3·2SiO2·2H2O(s) → Al2O32SiO2(s) + 2H2O(g) Δh2 1158a
Illite/Muscovite mica r3 (1/2)K2O·(3/2)Al2O3·3SiO2H2O(s) → (1/2)K2O·(1/2)Al2O3·3SiO2 (s) + Al2O3(s) + H2O(g) Δh3 374a
Chlorite r4 (Mg, Fe, Al)6(Si,Al)4O10(OH)8 → (Mg, Fe, Al)6(Si,Al)4O6(s) + 4H2O(g) Δh4 323a
Paragonite r5 (1/2)Na2O·(3/2)Al2O3·3SiO2·H2O(s) → (1/2)Na2O·(1/2)Al2O3·3SiO2 + Al2O3(s) + H2O(g) Δh5 236

Quartz inversion
Quartz inversion r6 α-SiO2(s) → β-SiO2(s) Δh6 11.7

Carbonate decomposition
Calcite r7 CaCO3(s) → CaO(s) + CO2(g) Δh7 1796
Dolomite r8 CaMg(CO3)2(s) → CaO(s) + MgO(s) + 2CO2(g) Δh8 1646

Crystallisation of calcium phases — formation of new crystalline phases

Anorthite r9 CaO(s) + SiO2(s) + CaOSiO2(s) → CaO·Al2O3·SiO2(s) Δh9 −46
Wollastonite r10 CaO(s) + SiO2(s) → CaO·SiO2(s) Δh10 −760
Enstatite r11 2MgO(s) + 2SiO2(s) → 2MgO·SiO2(s) Δh11 −655
Diopside r12 CaO(s) + MgO(s) + 2SiO2(s) → CaO·MgO·2SiO2(s) Δh12 −666
Akermanite r13 2CaO(s) + MgO(s) + 2SiO2(s) → 2CaO·MgO·2SiO2(s) Δh13 −636
Gehlenite r14 2CaO(s) + Al2O3(s) + SiO2(s) → 2CaO·Al2O3·SiO2(s) Δh14 −500
Mullite r15 Al2O3·2SiO2(s) → (1/3)3Al2O3·2SiO2(s) + (4/3)SiO2(s) Δh15 −280

Glassy phase formation by fusion of the crystalline phase

Potassium feldspar r16 K2O·Al2O3·6SiO2(s) → K2O·Al2O3·6SiO2(l) Δh16 207
Albite r17 Na2O·Al2O3·6SiO2(s) → K2O·Al2O3·6SiO2(l) Δh17 226
a
When the enthalpy of formation of the minerals was unavailable, values from the literature were used (Hirono and Tanikawa, 2011; Ratzenberger and Vogt, 1993; Vogt and Vogt,
2004; Vogt and Thomas, 2012).
32 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39

Table 7
Organic matter combustion considered during thermal treatment and combustion enthalpies.

Organic matter Organic matter combustion Value (kJ/kg component i)

Biomass (olive pit known as “orujillo”) CxHyOz (s) + O2(g) → aCO2(g) + bH2O(g) −16,663
Polyvinyl alcohol PVA (CH2CH(OH))n (l) + O2(g) → xCO2(g) + yH2O(g) −21,310
Polyethylene glycol PEG 200 H-(O-CH2CH2)4-OH (l) + 10O2(g) → 8CO2(g) + 9H2O(g) −26,753

To determine the mineralogical phases present in the fired ceramic
compositions, an unfired test piece of each type of composition,
40 mm in diameter and 6 mm thick, was formed by uniaxial pressure
in a laboratory press. The average dry bulk density of these pieces was
2000 ± 0.005 g/cm3. Each test piece was then fired to the composition's
peak temperature, detailed in Table 1. The heating rate was 5 °C/min.
For X-ray diffraction analysis, between 10 and 30 g of each unfired
composition was dried in an oven at 110 °C for at least 2 h. The same
quantity of fired sample was also weighed. Each unfired and fired com-
position was ground in a tungsten carbide ring mill to a particle size
below 100 μm.
3.1.4. Material characterisation results
The seven studied traditional ceramic products had different miner-
alogical compositions, which determined the physico-chemical trans-
formations that developed during the firing cycle and, consequently,
product end properties. The chemical and mineralogical composition
of the seven test compositions are detailed in Table 2 and Table 3,
respectively. The sulphur content was negligible in all compositions.
The firing behaviour of the bodies differed as a function of the com-
position, formulated to obtain the targeted end properties (porosity,
mechanical strength, etc.) (Venturelli and Paganelli, 2007). Composi-
tions RBT and PCT, used to manufacture red-body stoneware tile and
porcelain tile, respectively, as well as the porcelain tableware composi-
tion (PTW), formed liquid phase at about 900 °C, causing body porosity
progressively to decrease as the temperature rose (Orts et al., 1993). On
the other hand, compositions RWT, WWT, WBR, and RFT exhibited
quite a high, steady porosity in the range of test temperatures, owing
to the crystallisation of calcium silicates and aluminosilicates (gehlenite,
anorthite, and wollastonite) from the reaction of CaO, stemming from
calcite and dolomite decomposition, with Al2O3 and SiO2 from clay
mineral dehydroxylation (Cultrone et al., 2001; González-García et al.,
1990; Peters and Iberg, 1978; Riccardi et al., 1999). The mineralogical
composition of the fired samples is detailed in Table 4.

3
2
Cp (J/g K)

3 4
2

1
1
6

-1
5

-2
0 200 400 600 800 1000 1200
Temperature (ºC)
Cp·ΔT
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
40–315 1 1. Loss of moisture, decomposition of hydroxides and organic matter 31
2. Clay mineral dehydroxylation
315–625 2, 3 209
3. Quartz α β transformation
625–800 4 4. Carbonate decomposition 37
5. Crystallisation
800–1143 5, 6 18*
6. Partial fusion of the material
TOTAL 295
* Although crystallisation is an exothermic process and the fusion of the material is endothermic, the two processes have been grouped together and the total enthalpy is shown.

Fig. 2. DSC curve and physico-chemical transformations identified in the RBT composition.
 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39 33

3.2. Energy required to fire the studied ceramic materials
All the thermal energy that reaches a ceramic composition is used in
raising its temperature and in developing the different physico-
chemical transformations of the material in the firing process. This
heat can be calculated from the following equation, in which the first
term of the equation corresponds to the energy needed to raise the
composition temperature, and the second term is the total enthalpy of
the chemical reactions.
ZT f
QC ¼ mc  Cpc þ ΔH R
Ti
where:
– mc is the mass flow (kg/s) (in dynamic systems) or mass (kg)
(in static systems) of the ceramic composition that is conveyed
through or is stacked in the kiln, respectively.
– Cpc is the specific heat of the ceramic composition (J/kg K), which
varies as a function of the composition and temperature at each
moment.
– Tf and Ti are the peak firing temperature and the kiln entrance
temperature (K), respectively.
– ΔHR is the enthalpy of the physico-chemical reactions involved
(kJ/kg).
In this study, the enthalpy of reaction was determined in two ways:
by an experimental method (method 1) using differential scanning
calorimetry (DSC) and by an analytical method (method 2), performing
a calculation based on the mineralogical composition of the ceramic
body. The information required to calculate the heat of reaction in
each method is summarised in Table 5.
The information in Table 5 is set out in further detail below.
3.2.1. Experimental method
The enthalpy of reaction of the composition was determined in the
laboratory using a Netzsch model STA 449C DSC instrument. The instru-
ment enables a product's heat per unit mass in the firing process to be
quantified throughout a thermal cycle (Hatakeyama and Liu, 2000)
1 and allows measurement up to temperatures of 1350 °C, with an error
of 5–10%. The instrument was coupled to a thermogravimetric analysis
unit, thus simultaneously recording mass loss as a function of tempera-
ture. A description follows of the method used to determine the heat of
reaction.
A sample of the composition to be analysed and a standard sam-
ple were placed in a kiln and subjected to a controlled thermal
cycle. The standard was a sapphire disk of known thickness and cal-
orific capacity, which underwent no significant physico-chemical
transformation from an energy viewpoint in the test temperature
range.
The sample to be analysed needed to be in the form of a homoge-
neous fine powder with a particle size below 100 μm. To obtain the
test powders, the bulk samples were dried in an oven at 110 °C for at
least 2 h and then ground in a tungsten carbide ring mill. The thermal
treatments were carried out in platinum crucibles with their respective
covers, in a dynamic atmosphere of air and protective argon gas.

2
3
Cp (J/g K)

2
3
1
1

6
0

4
-1

5
-2
0 200 400 600 800 1000 1200
Temperature (ºC)

Cp·ΔΤ
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
1. Loss of moisture and/or adsorbed water, hydroxide
40–325 1 30
decomposition, and loss of water bound in clay structures
2. Clay mineral dehydroxylation
325–700 2, 3 264
3. Quartz α β transformation
900–1125 4, 5 4, 5. Crystallisation -24
1125–1190 6 6. Sintering start 32
TOTAL 303

Fig. 3. DSC curve and physico-chemical transformations identified in the PCT composition.
34 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39

In each DSC test, the temperature was raised from ambient temper-
ature to the composition peak firing temperature, using a 4-step proce-
dure: 1) from room temperature to 40 °C (v = 3 °C/min); 2) isotherm
at 40 °C (10 min); 3) from 40 °C to the peak temperature for each
composition detailed in Table 1 (v = 5 °C/min.); and 4) isotherm at
peak temperature (5 min).
The mass and temperature of the sample and of the standard, as well
as the required heat input to the sample to keep it at the same temper-
ature as the standard, were determined during the thermal cycle. This
information, together with a previous calibration that provided the co-
efficients of heat transmission in the firing chamber, enabled the heat
flux to the sample and from the sample to the kiln to be determined.
The test yielded the following information:
– Composition calorific capacity as a function of temperature. Inte-
grating this curve in each temperature range and multiplying
by the test mass enabled the heat of reaction in each range to be
obtained.
– Composition heat flux as a function of temperature. This curve
showed the heat exchange of the sample with the ambient. That is,
whether the chemical reactions that took place were endothermic
or exothermic.
– Composition mass loss as a function of temperature.
associated with each reaction, the heat released or absorbed, as well
as the approximate start and end of each reaction, to be identified. The
sum of the heat of reaction of the identified chemical reactions provided
the overall reaction enthalpy (ΔHR).
3.2.2. Analytical method
The heat of reaction was calculated by the analytical method accord-
ing to the following steps:
1. Determination and quantification of the mineralogical phases present
in the starting ceramic composition and in the final fired piece.
2. Identification of the most significant reactions during thermal treat-
ment of the composition from an energy viewpoint. These reactions
are shown in Table 6.
3. Calculation of the enthalpy of reaction of each transformation from
the enthalpy of formation of its products and reagents. These values
were either drawn from thermodynamic databases (Holland and
Powell, 2011) or, when unavailable, from the literature (Hirono and
Tanikawa, 2011; Ratzenberger and Vogt, 1993; Vogt and Vogt, 2004;
Vogt and Thomas, 2012).
4. Calculation, from the fractions by weight of each compositional
constituent, of the energy that each reaction contributed. The
sum of these contributions yielded the total heat of reaction of
the composition.

The joint interpretation of the three curves allowed the chemical After the mineralogical phases in the unfired and fired compositions
reactions of the composition during the thermal cycle, the mass loss had been determined and quantified, taking into account the physico-

3 4
Cp (J/g K)

2
2 3

1
1

0
5

-1

-2
0 200 400 600 800 1000 1200
Temperature (ºC)

Cp·ΔT
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
1. Loss of free and/or adsorbed water and water bound in clay
40–300 1 28
structures
2. Clay mineral dehydroxylation
300–680 2, 3 286
3. Quartz α β transformation
680–780 4 4. Carbonate decomposition 127
780–1126 5 5. Crystallisation -25
TOTAL 416

Fig. 4. DSC curve and physico-chemical transformations identified in the RWT composition.
 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39 35

chemical reactions found in the literature review (summarised in
Table 6), the most significant foreseeable reactions during firing, from
an energy viewpoint, were respectively assigned to each composition.
This method also takes into account the crystallisation of calcium
phases (exothermic process) and glassy phase formation in the
sintering process (endothermic process). These processes are complex
and, besides depending on the composition, they also depend on peak
temperature and residence time at peak temperature. Owing to their
complexity and relatively low energy value in the overall process, the
calculation of the energy contribution of glassy phase fusion was limited
to the energy calculation of the fusion of the fluxes, without taking into
account the other compositional components in the glassy phase.
However, in future studies it will be attempted to define these high-
temperature processes more extensively in the calculation method.
According to the principles of thermodynamics, enthalpy is a
function of state. Consequently, the standard enthalpy of reaction
(ΔH0R) can be calculated by subtracting the sum of the standard
enthalpies of formation of the reagents from the sum of the standard
enthalpies of formation of the products, as shown in the following
equation.
X X
0 0 0
ΔH R ¼ Δh f ðproductsÞ− Δh f ðreagentsÞ
The values of the enthalpy of formation of the mineralogical phases
used in the calculations were obtained from a thermodynamic database
of minerals prepared by Holland and Powell (2011), while the enthalpy
of formation of metakaolinite was obtained from various literature
sources (Schieltz and Soliman, 1964; Vaughan, 1955). When the enthal-
py of formation of the minerals was unavailable, such as that of the illite
and chlorite clays, values from the literature were used (Hirono and
Tanikawa, 2011; Ratzenberger and Vogt, 1993; Vogt and Vogt, 2004;
Vogt and Thomas, 2012). The use of these enthalpy formation values,
together with Eq. (2), enabled the enthalpies of reaction shown in
Table 6 to be calculated.
The values of the enthalpy of organic matter combustion used in the
calculations were obtained from various literature sources (Walters
et al., 2000; Retrieved July 10, 2014). These values are shown in Table 7.
Once the enthalpies of reaction had been obtained from the
enthalpies of formation of the components in each composition, the
total enthalpy of reaction of the ceramic composition was calculated,
this being the sum of the contribution of each reaction that developed.
The fractions by weight of each component in the composition
enabled the energy that each reaction contributed to be calculated.
The total enthalpy of reaction of the ceramic composition, applying
the values shown in Table 6, was calculated from Eq. (3):
X
ΔH R ¼ mi  Δhi 3
2
where mi is key component mass (kg) and Δhi its enthalpy of reaction
(kJ/kg).
To apply the analytical method, the key component was considered
to be the mineralogical phase identified in the mineralogical analysis of

2
3
4
Cp (J/g K)

3
2

1
1

0
5

-1

-2
0 200 400 600 800 1000 1200
Temperature (ºC)

Cp·ΔT
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
1. Loss of free and/or adsorbed water and water bound in clay
40–320 1 30
structures
2. Clay mineral dehydroxylation
300–630 2, 3 248
3. Quartz α β transformation
630–790 4 4. Carbonate decomposition 171
790 1140 5 5. Crystallisation -4
TOTAL 445

Fig. 5. DSC curve and physico-chemical transformations identified in the WWT composition.
36 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39

the starting mixture (reactions r1 to r8) or the new mineralogical phase
that formed (reactions r9 to r15). In reactions r16 and r17, the component
content was considered to be the percent weight difference between
the potassium feldspar or albite present in the starting and in the final
mixture. A degree of conversion equal to unity was assumed for all
reactions.
4. Results and discussion
The heat of reaction required for the physico-chemical transforma-
tions of the seven test compositions was determined according to the
two described methods. To perform the calculations, the compositions
were assumed to be dry, as the tests were conducted using dry test
pieces. The energy calculations for each studied composition are set
out below, and the results obtained with the two methods are
compared.
The physical and chemical transformations that developed during
firing in the studied compositions were identified from the curves
shown below, obtained by DSC analysis. The curves relate the calorific
capacity of the material to the temperature, and the numbers identify
the peaks observed. The transformations observed in the seven studied
compositions, as well as the temperature range and energy exchanged
in each transformation, are broken down in Figs. 2 to 8.
Table 8 summarises the heat of reaction of the seven studied compo-
sitions determined by the experimental method, the values obtained
using the proposed analytical method, and the difference in the results
obtained by the two methods.
The following observations may be drawn from the results obtained
in Figs. 2 to 8 and Table 8:
• The figures show that, in all compositions, several transformations
took place in the material when the unfired products were heated
from ambient temperature to about 800 °C. The most important trans-
formations were removal of free water, organic matter combustion,
allotropic transformation of α-quartz to β-quartz, loss of OH− groups
in the clays, and carbonate decomposition when the composition
contained carbonates.
• Raising the temperature from about 800 °C to the programmed peak
temperature (between 850 °C and 1200 °C, depending on the compo-
sition) caused the physico-chemical transformations to develop that
determined end product properties and led to the most important
dimensional changes (shrinkage) in the low porosity products (stone-
ware and porcelain ware) and to the formation of stable crystalline
phases (calcium silicates) in the porous products (earthenware wall
tiles and structural ceramics).
• From an energy viewpoint, clay mineral dehydroxylation and carbon-
ate decomposition were the principal endothermic reactions, while
organic matter combustion was the major exothermic process.
The total heat of reaction therefore depended mainly on these
transformations.
• Table 8 shows that the difference between the heat of reaction values
obtained by the analytical method and those obtained by the experi-
mental method for all compositions was quite small. The greatest
difference was observed with compositions WBR and PTW, though

6
5

3
Cp (J/g K)

2 4
3
1
1
2

-1

-2
0 200 400 600 800 1000 1200
Temperature (ºC)
Cp·ΔT
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
1. Loss of free and/or adsorbed water and water bound in
40–220 1 10
clay structures, decomposition of hydroxides
220 450 2 2. Organic matter combustion -82
3. Clay mineral dehydroxylation
450 630 3, 4 86
4. Quartz α β transformation
630–850 5 5. Carbonate decomposition and heating to end temperature 306
TOTAL 320

Fig. 6. DSC curve and physico-chemical transformations identified in the WBR composition.
 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39 37

6
4

3
Cp (J/g K)

2
3
2
1
1

0
6
5
-1

-2
0 200 400 600 800 1000 1200
Temperature (ºC)

Cp·ΔT
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
1. Loss of free and/or adsorbed water and water bound in
40–275 1 clay structures, decomposition of hydroxides and organic 33
matter
2. Clay mineral dehydroxylation
275–620 2, 3 131
3. Quartz α β transformation
620–760 4 4. Carbonate decomposition 282
760 1019 5, 6 5, 6. Crystallisation -30
TOTAL 416

Fig. 7. DSC curve and physico-chemical transformations identified in the RFT composition.

the value remained within the allowable experimental uncertainty.
The values obtained by the analytical method thus exhibited good
agreement with the experimental data.
• Composition organic matter content needed to be taken into account
in evaluating the heat of reaction. In the PTW composition, for
instance, the organic additives used as binder and plasticiser signifi-
cantly modified the heat of reaction. It may further be noted that, in
the WBR composition, the organic matter was specifically introduced
as fuel to reduce kiln energy consumption.
• Comparison of the energy consumption per unit mass shows that
the WWT, RWT and RFT compositions consumed the most energy,
followed by the PTW composition, the former owing to their larger
carbonate content and the latter to its high kaolinite content.
Although the WBR composition had the highest carbonate content,
its energy consumption was intermediate. This was presumably related
to the added organic matter and relatively low firing temperature. The
PCT and RBT compositions consumed less specific energy, mainly
because of their low carbonate content.
4.1. Final comparative analysis
The results of the two studied methods (Table 8) have been plot-
ted for visual comparison in Fig. 9, in which the enthalpies of reaction
of the seven test compositions calculated by the analytical method
are plotted versus their experimentally determined enthalpies of
reaction.
Good agreement between both sets of values can be observed, a
linear relationship being obtained with a coefficient of regression of
0.8905 and slope close to 1.
These results show that the proposed analytical method was appro-
priate for estimating the heat of transformation of the tested ceramic
compositions, in the studied heat of reaction range.
5. Conclusions
The following conclusions were drawn from the study:
○ Differential scanning calorimetry (DSC) enabled the energy needed
to fire seven typical body compositions used in traditional ceramic
manufacture to be determined, the main transformations that took
place during the firing cycle to be identified, and the energy involved
in each transformation to be quantified.
○ An analytical methodology was developed to determine the heat of
reaction of traditional ceramic compositions, based on: 1) identifica-
tion of the main physico-chemical transformations from the analysis
of the unfired and fired mineral content; 2) the energy involved in
these transformations, obtained from the literature and thermody-
namic databases; and 3) the presence and type of organic matter.
The results obtained exhibited good agreement with the experimen-
tal data obtained by DSC analysis.
○ The results show that the heat of reaction in the studied traditional
ceramic compositions is closely linked to the carbonate and clay
38 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39

6
3

3
Cp (J/g K)

4
2

1
1

2
0 5

-1

-2
0 200 400 600 800 1000 1200
Temperature (ºC)

Cp·ΔT
Temperature range (ºC) Peaks Interpretation
(kJ/kg)
1. Loss of free and/or adsorbed water and water bound in
40–350 1, 2 clay structures, decomposition of hydroxides 15
2. Organic matter combustion
3. Clay mineral dehydroxylation
350–650 3, 4 412
4. Quartz α β transformation
950–1000 5 5. Crystallisation -23
TOTAL 404

Fig. 8. DSC curve and physico-chemical transformations identified in the PTW composition.

mineral (especially kaolinite) content. On the other hand, it was
verified that even low quantities of organic matter significantly
influence the final heat of reaction, as organic matter combustion
was highly exothermic. Consequently, the organic matter content
in such compositions, including organic additives such as binders
and plasticisers, needs to be quantified in order to properly estimate
the heat of reaction.
○ The results obtained and the methodology developed in this study
allow the energy involved in the firing of traditional ceramic compo-
sitions to be estimated. However, this estimation is not limited to the
studied compositions, as other reactions can also be taken into ac-
count, depending on the raw material composition. The developed
analytical method can be used in simulation and optimisation stud-
ies on raw material formulations, in industrial energy studies (e.g. in
drawing up energy balances or estimating kiln efficiency), as well as
in environmental studies (e.g. in life cycle assessments or in measur-
ing carbon dioxide emissions or carbon footprints).
Acknowledgements
This study was funded by the Valencian Institute of Business Com-
petitiveness (IVACE) in the Competitiveness Improvement Programme
through project IMAMCA/2014.

Table 8
Heat of reaction of the studied ceramic compositions.

Composition Heat of reaction

Experimental method (kJ/kg unfired tile) Analytical method (kJ/kg unfired tile) Difference (%)

Red-body stoneware tile RBT 295 ± 30 298 ± 35 1.0

Porcelain tile PCT 303 ± 30 320 ± 25 5.6
Red-body earthenware wall tile RWT 417 ± 42 433 ± 35 3.8
White-body earthenware wall tile WWT 444 ± 44 418 ± 36 5.9
White brick WBR 320 ± 32 341 ± 42 6.5
Roof tile RFT 416 ± 42 411 ± 62 1.2
Porcelain tableware PTW 404 ± 40 430 ± 30 6.6
 S. Ferrer et al. / Applied Clay Science 108 (2015) 28–39
500
450
ΔHR Analytical (kJ/kg)
400
350
WBR
PCT
300
RBT
250
250 300
ΔHR Experimental (kJ/kg)
Fig. 9. Experimental heat of reaction data versus the values calculated by the analytical method.
References
Agrafiotis, C., Tsoutsos, T., 2001. Energy saving technologies in the European ceramic
sector: a systematic review. Appl. Therm. Eng. 21 (12), 131–1249.
Bernardo, E., De Lazzari, M., Colombo, P., Saburit, A., García-Ten, J., 2010. Lightweight
porcelain stoneware by engineered CeO2 addition. Adv. Eng. Mater. 12 (1–2), 65–70.
Bovea, M.D., Díaz-Albo, E., Gallardo, A., Colomer, F.J., Serrano, J., 2010. Environmental
performance of ceramic tiles: improvement proposals. Mater. Des. 31, 35–41.
Brindley, G.W., Brown, G., 1984. Crystal Structures of Clay Minerals and Their X-ray
Identification. Mineralogical Society, London.
Castelein, O., Soulestin, B., Bonnet, J.P., Blanchart, P., 2001. The influence of heating rate on
the thermal behaviour and mullite formation from a kaolin raw material. Ceram. Int.
27, 517–522.
Cultrone, G., Rodriguez-Navarro, C., Sebastian, E., Cazalla, O., De la Torre, M.J., 2001.
Carbonate and silicate phase reactions during ceramic firing. Eur. J. Mineral. 13,
621–634.
Eliche, D., Martinez, C., Martínez, M.L., Cotes, M.T., Pérez, L., Cruz, N., et al., 2011. The use of
different forms of waste in the manufacture of ceramic tricks. Appl. Clay Sci. 52,
270–276.
European Commission, 2007. Reference Document on Best Available Techniques in
the Ceramic Manufacturing Industry (Available at: http://eippcb.jrc.ec.europa.eu/
reference/).
González-García, F., Romero-Acosta, V., García-Ramos, G., González-Rodríguez, M., 1990.
Firing transformations of mixtures of clays containing illite, kaolinite and calcium
carbonate used by ornamental tile industries. Appl. Clay Sci. 5 (4), 361–375.
Hatakeyama, T., Liu, Zhenhai (Eds.), 2000. Handbook of Thermal Analysis. John Wiley &
sons, Chichester.
Hirono, T., Tanikawa, W., 1 July 2011. Implications of the thermal properties and kinetic
parameters of dehydroxylation of mica minerals for fault weakening, frictional
heating, and earthquake energetics. Earth Planet. Sci. Lett. 307 (1–2), 161–172.
Holland, T.J.B., Powell, R., 2011. An improved and extended internally consistent thermo-
dynamic dataset for phases of petrological interest, involving a new equation of state
for solids. J. Metamorph. Geol. 29, 333–383.
Klein, C., Hurlbut Jr., C.S., 1985. Manual of Mineralogy, After James D. Dana, 21st Edition,
Revised. 21st edition. John Wiley & Sons, New York (681 pp.).
Macgee, A.E., 1926. The heat required to fire ceramic bodies. J. Am. Ceram. Soc. 9, 206–247.
Madivate, C.M., Malate, A.M., Verryn, S., Loubser, M., 2004. Energy requirement for firing
porcelain. Bull. Chem. Soc. Ethiop. 18 (1), 73–80.
Mezquita, A., Monfort, E., Zaera, V., 2009. Ceramic tiles manufacturing and emission
trading scheme: reduction of CO2 emissions, European benchmarking. Bol. Soc. Esp.
Ceram. Vidrio 48 (4), 211–222.
39
RWT
PTW
WWT
RFT
Analytical = 1.0166·Experimental
R2 = 0.8905
350 400 450 500
Mezquita, A., Boix, J., Monfort, E., Mallol, G., 2014. Energy saving in ceramic tile kilns:
cooling gas heat recovery. Appl. Therm. Eng. 65 (4), 102–110.
Monfort, E., Mezquita, A., Granell, R., Vaquer, E., Escrig, A., Miralles, A., Zaera, V., 2010.
Analysis of energy consumption and carbon dioxide emissions in ceramic tile
manufacture. Bol. Soc. Esp. Ceram. Vidr. 49 (4), 303–310.
Njoya, D., Hajjaji, M., Njopwouo, D., 2012. Effects of some processing factors on technical
properties of a clay-based ceramic material. Appl. Clay Sci. 65–66, 106–113.
Orts, M.J., Escardino, A., Amorós, J.L., Negre, F., 1993. Microstructural changes during the
firing of stoneware floor tiles. Appl. Clay Sci. 8 (2/3), 193–205.
Pennisi, L., 1991. In ceramics and glasses. Engineering Materials Handbook vol. 4. ASM
International, Materials Park, OH, pp. 255–259.
Peters, T., Iberg, R., 1978. Mineralogical changes during firing of calcium-rich brick clays.
Am. Ceram. Soc. Bull. 57, 504.
Plante, A., Fernández, J.M., Leifeld, J., 2009. Application of thermal analysis techniques in
soil science. Geoderma 153, 1–10.
Ratzenberger, H., Vogt, S., 1993. Possibilities for the Prediction of Shaping, Drying and
Firing Behavior of Heavy Clays and Production Bodies ZI — Annual 1993. pp. 70–111.
Retrieved January 16, 2014, from. http://www.icdd.com/.
Retrieved July 10, 2014, from. http://www.dow.com/polyglycols/polyethylene/products/
carbowaxp.htm.
Riccardi, M.P., Messiga, B., Duninuco, P., 1999. An approach to the dynamics of clay firing.
Appl. Clay Sci. 15 (3–4), 393–409.
Sánchez, E., Orts, M.J., García-Ten, J., Cantavella, V., 2001. Porcelain tile composition effect
on phase formation and end products. Am. Ceram. Soc. Bull. 80 (6), 43–49.
Schieltz, N.C., Soliman, M.R., 1964. Thermodynamics of the various high-temperature
transformations of kaolinite in clays and clay minerals. Proceedings of the Thirteenth
National Conference. Pergamon, London, pp. 419–428.
Simpson, H.E., 1927. The heat required to fire ceramic bodies. J. Am. Ceram. Soc. 10,
897–918.
Vaughan, F., 1955. Energy changes when kaolin minerals are heated. Clay Miner. Bull. 2,
265–274.
Venturelli, C., Paganelli, M., 2007. Sintering Behaviour of Clays for the Production of
Ceramics (cfi/Ber. DKG 84, No. 5).
Vogt, S., Thomas, R., 2012. On the preparation of energy balances for brick and tile-making
plants (part 2). ZI Int. 64 (10), 12–27.
Vogt, S., Vogt, R., 2004. Relationship between minerals and the industrial manufacturing
properties of natural clay deposits and the clay bodies produced from them for the
heavy clay industry (part 2). ZI — Annual 2004pp. 78–103.
Walters, R.N., Hackett, S.M., Lyon, R.E., 2000. Heats of combustion of high temperature
polymers. Fire Mater. 24, 245–252.
Zevin, L.S., Kimmel, G., 1995. Quantitative X-ray Diffractometry. Springer, New York.